Several people cited the AHS-2 as a pseudo-RCT that supported veganism. There’s one commenter on LessWrong and two on EAForum
The LessWrong link here links to my comment even though I did not describe the AHS-2 as a “pseudo-RCT.” So I think your description of that comment is misleading.
Edit: it’s also misleading to say that “Outcomes for veganism are ~tied with pescetarianism in men, and are worse than everything except for omnivorism in women” in AHS-2, as I explained on LW.
For what it’s worth, I interpreted the original post as Elizabeth calling it a pseudo RCT, and separately saying that commenters cited it, without implying commenters called it a pseudo RCT
I understood ‘pseudo’ here to mean ‘semi’ not ‘fake’. So my interpretation of Elizabeth’s argument is ‘people point to this study as a sort-of-RCT but it really doesn’t resemble that’
On one hand, you did go out of your way to describe AHS-2 as “not a longitudinal RCT”. That’s a pretty good argument you didn’t represent it as an RCT.
On the other hand, you comment consists mostly of quotes from a comment that only makes sense if diet was an independent variable in the study. The post that comment cites does the same, more explicitly. I think “cited as a pseudo-RCT” is a fair way to describe those comments quickly.
But ultimately my point depends less on the word choice ‘pseudo-RCT’ than on how you used AHS-2 in the discussion. I asked you about concerns I had with the study twice, and wrote a full post you could have responded to. You also expressed an intention to provide more information on the RCTs you linked to, and haven’t.
If you have alternate phrasings you think capture all the information in the same space more accurately, I’m open to that. But I suspect we won’t be able to come to agreement on phrasing without diving into the real disagreements on content.
The LessWrong link here links to my comment even though I did not describe the AHS-2 as a “pseudo-RCT.” So I think your description of that comment is misleading.
Edit: it’s also misleading to say that “Outcomes for veganism are ~tied with pescetarianism in men, and are worse than everything except for omnivorism in women” in AHS-2, as I explained on LW.
For what it’s worth, I interpreted the original post as Elizabeth calling it a pseudo RCT, and separately saying that commenters cited it, without implying commenters called it a pseudo RCT
I understood ‘pseudo’ here to mean ‘semi’ not ‘fake’. So my interpretation of Elizabeth’s argument is ‘people point to this study as a sort-of-RCT but it really doesn’t resemble that’
On one hand, you did go out of your way to describe AHS-2 as “not a longitudinal RCT”. That’s a pretty good argument you didn’t represent it as an RCT.
On the other hand, you comment consists mostly of quotes from a comment that only makes sense if diet was an independent variable in the study. The post that comment cites does the same, more explicitly. I think “cited as a pseudo-RCT” is a fair way to describe those comments quickly.
But ultimately my point depends less on the word choice ‘pseudo-RCT’ than on how you used AHS-2 in the discussion. I asked you about concerns I had with the study twice, and wrote a full post you could have responded to. You also expressed an intention to provide more information on the RCTs you linked to, and haven’t.
If you have alternate phrasings you think capture all the information in the same space more accurately, I’m open to that. But I suspect we won’t be able to come to agreement on phrasing without diving into the real disagreements on content.
Hi—I replied to one of your questions on LW. Sorry for the delay, I’m quite swamped with work.